AI-based coronary plaque evaluations highlight elevated heart risks in women

She said this can add incremental value to improve risk stratification in women and men. Traditional, clinical risk scores have shown only moderate performance, especially in women, prompting the need for more refined tools. According to Feuchtner, AI QCT from Cleerly Health not only improves image analysis speed and accuracy but also identifies subtler plaque characteristics that are more predictive in female patients.

Notably, the study confirmed that non-calcified plaque, which was previously underestimated, serves as a powerful, independent predictor of adverse events in both sexes. Yet, when volumes are equal, women exhibit a steeper rise in risk.

"When you use a volumetric plaque increase, then these features indicate a higher relative risk in women," she said.

The initial data from CONFIRM2 presented at the Transcatheter Cardiovascular Therapeutics (TCT) 2024 meeting last fall  showed that the soft plaque is a very important metric that has been underestimated.

"It is well known high-risk plaque, vulnerable lipid-rich core plaque, is predictive for events. But the prevalence was very low. In our study, it was only 2.5%, which means only two or three out of a 100 women have vulnerable lipid-rich plaque, which cannot explain all the heart attacks. So noncalcified plaque volume was an independent predictor of events in both men and women. That was one of the interesting key findings," Feuchtner said.

AI greatly speeds detailed CCTA quantification and analysis

Beyond its diagnostic value, the technology holds significant clinical potential. AI-enabled analysis can be completed in under 10 minutes, which is far faster than manual caliper assessments that take up to an hour. She said they are so labour intensive, they are not done in clinical practice. But the AI now automates this and makes it accessible to wide clinical use. More importantly, she said it enables clinicians to monitor plaque progression or regression over time, so physicians can adjust drug therapies accordingly.

“This is the perfect technique to identify high-risk patients who would benefit from intensive therapies,” Feuchtner noted. “Non-calcified plaque can be treated and stabilized with statins or other agents, and the AI quantification lets us measure that impact longitudinally.”

Feuchtner emphasized that this is just the beginning. As the CONFIRM2 registry continues to grow, researchers aim to define plaque volume thresholds and explore additional features—up to 16 per patient—such as disease diffuseness and perivascular fat.

The future of CCTA in cardiac prevention and assessment

With its ability to enhance precision in risk stratification and treatment decisions, AI QCT is poised to change the standard of care in cardiovascular imaging—especially for women, whose unique biology has long been underserved by one-size-fits-all diagnostic tools.

"The impact of the trial and the results will be that we can improve risk stratification for men and women. Women are different and we have to be aware of that. Women have different thresholds. What the study shows is that at a certain plaque volume threshold, the risk in women increases sharply, while in man the curve is flat," Feuchtner explained.

She said CCTA AI plaque assessments could be a paradigm shift in how patients are evaluated and managed in the near future, but additional research is needed to develop more accurate risk assessments for the guidelines.

"There are important differences that we then will be looking in our final cohort to get more insights and other features of atherosclerosis. Not only the key metrics, but the AI technology enables use to analyze 16 different features of atherosclerosis that have yet to be explored in the future," Feuchtner said.

She said cardiac CT will play a bigger role in cardiac assessments, and is already having an impact to decrease the volumes for diagnostics catheterizations. She said this will be accelerated with new CT technology, including the use of photon-counting CT to enhance the detail seen in soft plaque imaging and the ability to see through metallic stents.

"New techniques will lead to a rise in patients referred to coronary CTA, which is already now the case," Feuchtner concluded.